CHAPTER 8
Cliffs, Quarries, Linear and Artificial Habitats
IF WE LOOK DOWN on the Lower Wye from a great height we see not only the coarse patchwork of fields, woods, heaths, lakes, golf courses, towns, villages and industrial estates, but also a variety of lines and small patches. Most lines are hedges and other artificial boundaries, such as lanes, motorways, railways, drainage ditches and power transmissions, but some are natural, notably rivers, streams, cliffs and scarps. If we then look at the detail, we see that woods are not homogeneous, but mosaics of different ages and kinds of tree cover, delimited by tracks and rides, and that fields contain wet hollows and disturbed ground around feeding troughs and gates. Focusing down further, we see, for example, that molehills diversify semi-natural grasslands and fallen logs diversify ground conditions in woodland. Conversely, if we stand back and look at the larger picture, we see large-scale patches, such as districts with concentrations of woods, or districts with large, rectangular fields that are clearly distinct from adjacent concentrations of small, irregular fields. Taken as a whole, therefore, the landscape has an extremely complex structure at various scales. This is the context in which wild species must survive and thrive.
This chapter complements the previous three by considering linear, small-patch and artificial habitats, a mixed bag that has one common feature: all the habitats are small or narrow, and thus strongly influenced by their surroundings. They are also foci of wildlife diversity, because they form edges between habitats and differ from their surroundings – how much wildlife would one find in arable districts if there were no hedges and road verges? Edges are particularly important for wildlife. A boundary between a wood and a field, for example, provides niches for species of shade, semi-shade and open conditions, and for species that need both the shelter of woods and feeding sites outside woods. Edges, especially hedges, are also thought to be the highways along which species move through the landscape, a point to which we will return at the end of the chapter.
The extent of these habitats is difficult to determine exactly, but estimates can be made. The length of field boundaries in an imaginary field system consisting of a grid of 1ha square fields would be 20km per km2, and if these boundaries were 2m wide, the boundaries would occupy some 4% of the land area. At the other end of the scale, if the fields extended to 20ha and the boundaries were 1.5m thick, the length of boundaries would be 4.5km per km2 and the land take would be 0.67%. These estimates correspond respectively to the small-field landscapes around the gorge and the most open arable landscapes of the Herefordshire lowlands, so an average density of boundaries is likely to be closer to a 10ha grid of boundaries 1.75m thick, which would occupy 1.2% of the land, and on this basis the 71,000ha of farmland in the Lower Wye contains about 4,600km of boundary habitats. If we add the 7.5% of the Lower Wye occupied by artificial habitats (Chapter 3) to the 1.2% occupied by boundaries, we find that artificial habitats and field boundaries occupy rather more land than the small fields and other non-woodland, semi-natural habitats combined.
CLIFFS
Natural cliffs form the main dramatic element in both the upper and lower gorge, the natural rugged features that contributed the sublime element to William Gilpin’s picturesque scenes. Emerging as pale scars from within the woods, the taller cliffs have never been heavily shaded by trees, yet they remain integral parts of ancient woodland (Fig. 139). If any places within the Lower Wye have remained in a natural state, the faces of the Seven Sisters, Coldwell Rocks, Wyndcliff, Shorncliff and their like appear to be as close to pristine as anywhere.
Cliffs generate a range of conditions. The tops of rock towers and bluffs have exposed, dry, skeletal soils on which the growth of trees and shrubs is repeatedly set back by heat and drought. Deer also use them as vantage points and reinforce the open condition by grazing and browsing. The upper faces of larger cliffs are also dry and open, for they remain only lightly shaded by overhanging trees and free from ground-water seepages. Conditions, however, differ between aspects: for example, the south-facing cliffs of the Doward receiving full sun are far more severe habitats than the perpetually shaded north-facing cliff of Lady Park Wood opposite. In contrast, the smaller cliffs and the lower levels of taller cliffs are
FIG 139. Some of the Seven Sisters rocks emerging from the ancient woods of the Great Doward. Although they are not as spectacular as longer cliffs, such as Wyndcliff and Ban-y-Gor rocks, these are botanically the most interesting.
shaded by trees growing from crevices and in soil at their foot. Irrespective of aspect, they tend to be moist habitats, patchily irrigated by seepages, and often covered by sheets of mature ivy or masses of bryophytes and ferns.
Conditions on the cliffs extend into the surrounding habitats. Immediately above, where soil is thin, dry and mixed with outcrops, tree growth is limited and shrub species are abundant. Above the Seven Sisters, such ground is occupied by a rich assemblage of wood-edge species, such as bloody crane’s-bill and madder. Laterally, the cliffs continue as very steep slopes with small outcrops and pockets of stony, very dry soils. Ordinary closed-canopy woodland can form on these sites, but, being vulnerable to drought and slope-induced disturbance, it is more likely to be open than woodland on mesic sites and flat ground. Below the cliffs we find boulders and screes, extremely irregular ground, with pockets of litter and mineral soil, where the ground is sheltered, shaded and often irrigated by seepages. Niches thus range from extremely dry on rocks to marshes on the strongest seepages. Cliff-margin conditions are actually widespread, for all steep woods have small outcrops with thin, dry soil. Thin, dry, rocky soils also develop on flat ground, especially over the limestone pavements on the Great Doward and in Tidenham Chase.
With such a diversity of extreme habitats in close proximity, the cliffs inevitably form centres of plant diversity. The grassland relicts on the most exposed towers and bluffs were described in Chapter 6, and the adjacent woodland has been described as type 2 in Chapter 5. Somewhat surprisingly, the exposed upper cliffs are less distinctive as habitats (Fig. 140): many faces are bare, and not just because they are used by rock climbers. Others are covered in thick fans of mature ivy, which exclude less robust species. Ledges and cracks support a scatter of common species, such as herb-robert, great lettuce, wall lettuce, dandelion and wood sage, but most of the distinctive species, such as red valerian and wallflower at Wintour’s Leap and the cabbages below Chepstow Castle, are, or may be, escapes from cultivation. Shaded cliffs support more bryophytes and ferns, including the pachyrachis form of maidenhair spleenwort. A few rare herbs have been recorded: mossy saxifrage, for example, was recorded on the Wyndcliff.
Tree and shrub species are markedly more diverse on and around the cliffs than in the surrounding woodland, reflecting the specialist niches, the absence of management for timber and the reduced competition offered by the main forest trees, whose growth is inhibited by seasonal drought. Thus, we constantly find species that are otherwise concentrated on edges and in recent openings, particularly dogwood, spindle, broom, privet, spurge-laurel, wayfaring-tree and several species of rose (small-leaved sweet-briar, small-flowered sweet-briar, sweet-briar, burnet rose, harsh downy-rose). Mezereon and bird cherry were formerly recorded from cliffs in the lower gorge, but have now vanished. Whitebeams and related species (Chapter 9) are particularly common around the cliffs, where their silvery foliage contrasts stunningly with the yew and beech in spring.
The ground vegetation in woodland around the cliffs is also notably diverse. True, the dominants are usually widespread species, such as dog’s mercury and hart’s-tongue, and common species that are strongly associated with dry, limestone soils, such as wall lettuce and early dog-violet, but there are also many others. Some are nationally uncommon, such as narrow-leaved bitter-cress, white and narrow-leaved helleborines, yellow bird’s-nest, ivy broomrape and serrated wintergreen. Some are near the northern or southern limits of their ranges, such as madder and mountain melick, respectively. Widespread species that always attract attention may be present, such as lily-of-the-valley, Solomon’s-seal and angular Solomon’s-seal.
One curious feature of the cliffs and their surroundings is the way that ordinary ecological distinctions tend to break down. Ecologists get used to the idea that particular species are associated with alkaline or acid soils, but around
FIG 140. Wintour’s Leap, a great rock face at the northern end of Pen Moel Rocks. Houses and gardens extend to the cliff edge above, allowing wallflowers to colonise the cliffs. Below, the tidal reach of the Wye passes below ancient woodland on the Lancaut peninsula, which is managed as a reserve by the Gloucestershire Wildlife Trust.
the cliffs the two groups tend to come together. Thus, wood sage is a characteristic species of both heathland and limestone outcrops. Wood fescue, a nationally rare woodland grass, is common on steep slopes on both the acid soils of the Hudnalls and the dry alkaline slopes above the Shorncliff. In lowland Britain, lily-of-the-valley is a species of woods on acid sands, but in the Lower Wye it grows on limestone. The same can be said of the trees: beech, sessile oak, small-leaved lime, holly and yew are all characteristic limestone species in the Lower Wye, but they are just as common on dry, acid sites.
The diversity of open-cliff vegetation has been sadly reduced in the last century or so. For example, hutchinsia has been lost from all but one of the five cliffs from which it has been recorded; hairy rock-cress from five of the eight; saw-wort from six of the eight; soft-leaved sedge from two of the eight, and mountain everlasting from Coldwell Rocks, its only cliff location. The endemic hawkweed Hieracium pachyphylloides epitomises the fate of cliff species. This glaucous-leaved microspecies with large, yellow heads was first noted in the late nineteenth century and has only ever been found in the Wye gorge, i.e. it is a Lower Wye endemic. Until 1910, it was regularly recorded at Piercefield Cliffs and Great Doward, but when Tim Rich and Jérôme Sawtschuk mounted a search, they finally found 114 plants in June 2006 in the third of its former locations, on the limestone exposures around Symonds Yat (Sawtschuk, 2006). There, the fate of the remaining three populations is in the hands of climbers, who are liable to damage what to them are insignificant plants, but who at least keep the faces open and thereby prevent the hawkweed from being shaded out.
Why has the cliff flora declined so much? The most likely explanation is the extinction of traditional woodland management (and thus more shade), combined possibly with less grazing from above the cliffs. As the images left by the Wye tourists indicate, the cliffs are less exposed now than they were two centuries ago, and it is on the remaining exposed cliffs that the flora survives. If this is right, and if the rare cliff species, such as bloody crane’s-bill and dwarf sedge, are relicts of prehistoric, more open conditions, then the woods have presumably been more open throughout history and much of prehistory than they are now. However, other factors may be at work: mysteriously, many shade species associated with limestone outcrops have also declined, not just the collectors’ targets, such as mezereon and fly orchid, but also yellow bird’s-nest (lost from four of nine cliffs) and stone bramble (lost from all five sites).
WALLS AND BUILDINGS
When Robert Bloomfield visited Chepstow Castle in 1811, he remarked in his diary about Marten’s Tower that
it appears unaccountable to me, how even by the lapse of ages, nutriment enough can be found for shrubs of so large a growth as are flourishing between the outer and inner ramparts of these towers, and on the top of the wall of course. This gangway, once the place of the defenders of the fortress and its sentinels, is now absolutely wild: a mixture of bramble, hazel, ash, beech, and fruit-trees, from twenty to thirty feet high at least.
Likewise, eighteenth- and nineteenth-century prints of Tintern Abbey consistently show the ruins festooned in vegetation, most of it very large growths of ivy. Later, Clark’s (c.1862) tourist guide indicated that a visit to the abbey in the nineteenth century must have been a botanical as well as a cultural experience, for he mentioned that the stonework was covered in a luxuriant growth of mosses, lichens, ferns, pennyworts (navelwort), wallflowers and snapdragons. When Francis Kilvert visited the abbey in June 1875, he rather daringly climbed to the top of the walls, which he found to be ‘adorned with a perfect wild-flower garden of scarlet poppies, white roses, yellow stonecrop, and purple mallow’. This vegetation may even have colonised while the great monastery was still in use, for about 1490, Tintern was ‘threatened in ruin in its walls, roofs, houses, and granges [by] the passage of years and [by] negligence and incompetence.’ (Williams, 2001, p. 83). Now the masonry is as clean as a whistle, cleaned and stabilised in the early twentieth century, leaving only a drapery of scaffolding.
Two types of wall should be recognised: free-standing walls with two sides, and retaining walls that hold back banks of soil or rock. The latter are more like natural outcrops, stony woodland and the banks of lanes – dry on the surface, but supplying moisture and deep rooting opportunities behind. They support abundant growths of bryophytes, ivy, hart’s-tongue, male, buckler and other ferns, especially in the sandstone districts around the gorge, but otherwise have few distinctive features.
Free-standing walls, on the other hand, are extreme habitats, seasonally parched (especially on the south-facing side), lacking in soil, and offering limited rooting depth, just like the upper faces of tall cliffs, clean quarry faces, free-standing boulders and tombstones. Wall construction has a substantial influence on the vegetation. Most are held together with lime-rich mortar, so the rooting medium is generally alkaline, though dry-stone walls made from sandstone and conglomerate must be extremely acid. Narrow walls built with the minimum of material are harsh habitats indeed, but the very wide dry-stone walls that are so common in the sandstone and conglomerate districts generally have a core of rubble and humus into which deep rooting is possible. Most walls have been built with local rock, but walls in the sandstone districts were often made from imported limestone, and these can be identified by their cover of rustyback and wall-rue.
The conglomerate walls on the Hudnalls demonstrate a continuum from tall free-standing walls to natural boulder fields and outcrops. Formed around 1800, most are now overgrown with woodland and covered in bryophytes. While a few were carefully constructed, most were little more than stones and boulders heaped into lines and given a crude facing of carefully laid stones. Some seem to be the residue of hedges into which stones were thrown, others were just clearance cairns, and many have simply collapsed since the fields they enclosed were abandoned. Ecologically, the whole collection merges with the natural boulder fields in adjacent ancient woodland. A different kind of continuum is symbolised by Chepstow Castle, rising without a break from limestone cliffs over the Wye, and by the wild cabbages that grow indiscriminately on both walls and cliff.
The distribution of walls as field boundaries is patchy. Hedges are the rule in most of Herefordshire and much of the plateau land either side of the gorge, but walls are common in and near the gorge, especially in the vicinity of the puddingstone outcrop: indeed, the density of walls in parts of St Briavels Common must be one of the highest in Britain, and has recently been estimated to exceed 15km per km2 (Anon, 2005). Some half-hearted walls on the old commons appear never to have been much more than roughly stacked lines of rocks, but as trees and shrubs have colonised, these have changed into hedges over the years.
Walls age as habitats. Lichens and mosses grow first on bare stonework (Fig. 141), as they do on tombstones, but after a few years of growth and decay their humus provides both a rooting medium for flowering plants and a measure of buffering against drought. Given time, as the walls of Tintern Abbey once demonstrated, even free-standing walls become clothed in vegetation. Even shrubs and trees, such as ash, wych elm, elder and roses, germinate high on the masonry of ruined castles, though they rarely prosper, and they destroy the wall if they do. Ivy is particularly partial to both rock faces and walls, forming great fans of foliage on the larger faces, but it roots in the ground below and is no more dependent on walls than it is on trees. It ramifies through the structure, develops dense heads of mature foliage and, as the stems expand with age and the branches wave in the breeze, it eases the stones apart and eventually brings
FIG 141. Cladonia rangiformis forming a lichen heath on a 200-year-old wall of Quartz Conglomerate in the Hudnalls. This species is scarcely found locally in natural habitats, but may have been common on the boulders that were once strewn over the Hudnalls wood-pasture, the source of and reason for the walls.
about the collapse of the wall. Older walls tend to be richer communities: many species, such as yellow corydalis (Fig. 142), are especially associated with old walls. Latterly, however, wall floras may have been somewhat impoverished by greater attention to wall maintenance, and not just at Tintern Abbey and local castles, where the walls are now clean. The rare flattened meadow-grass has evidently gone from the walls of Chepstow, though it is still on the Seven Sisters.
The distinctive and common wall flora includes biting stonecrop and several other related species, each with small, rounded, succulent leaves that are clearly adapted, like cacti, to dry habitats. Another relative, the house-leek, used to be frequent on walls and roofs in the Dean. Limestone walls are distinctively clothed in red valerian, ivy-leaved toadflax, common whitlowgrass, pellitory-on-the-wall, and mixtures of small ferns, notably the silvery-green rustyback, wall-rue, maidenhair spleenwort and black spleenwort. Another common limestone fern is hart’s-tongue, but this requires more moisture than the smaller ferns.
Other plants are more catholic in their choice of substrate. The most distinctive is navelwort, with its rounded leaves about the size of an old penny, which spreads fairly rapidly, but is said to be more associated with old walls. Another common species on conglomerate walls is shining crane’s-bill, which is often accompanied by its relative, herb-robert. Polypody and the slightly larger intermediate polypody are widespread on the more acid walls, and may come to dominate shaded wall tops. Other fairly widespread species include wall lettuce, thale cress, and three small grasses, red fescue, silver hair-grass and early hair-grass. In practice, many walls are shaded and filled with litter and bryophytes, and these are likely to harbour common species, such as the brambles that find a multitude of crevices for their tip-rooting canes. Mysteriously, one wide wall that we had recently cleared of thick ivy became a temporary arable field, for wheat sprang up on the top, and this several kilometres from the nearest wheat field.
Walls are manifestly artificial habitats, so perhaps it is understandable that many of the most distinctive and colourful species are aliens, introduced mainly from hotter and seasonally dry Mediterranean or continental climates. Some, such as greater celandine and the medicinal herb feverfew, are pre-Norman introductions, and others escaped in the Middle Ages, for example wallflower and garden parsley. Most, however, are post-medieval escapees, and these include some of the commonest species of limestone walls, including red valerian and ivy-leaved toadflax. Several attractive garden plants have maintained self-sustaining populations on Lower Wye walls for a while, such as pinks, fairy foxglove and fox-and-cubs, but, like most escapees, they have not got very far. They may yet spread, for a slow start does not preclude eventual success: thus
FIG 142. Yellow corydalis on a limestone wall in Chepstow. This, like ivy-leaved toadflax, is an ancient colonist of walls that has scarcely spread to natural limestone outcrops. Red valerian on the other hand is now very widespread in natural and artificial habitats.
red valerian in 1889 did ‘not appear to have established itself anywhere in Herefordshire as a denizen, but is often found on the walls of old gardens’ (Purchas & Ley, 1889), but it is now common even in natural habitats. Wall whitlowgrass, a native of the Cotswolds, survived for many years as an introduced species of old walls in Clearwell and Newland. Likewise, the cabbages on Chepstow Castle could be either a native population or escapees from cultivation. Trivial Pursuit enthusiasts may like to know that, before recent nomenclatural changes, virtually all the common introduced species of Lower Wye walls had scientific names starting with the letter C.
Otherwise, the plants of walls come from a restricted range of more natural habitats. Several grow naturally on rock faces, including navelwort, biting stonecrop and the small ferns. More are characteristic of steep and rocky woods, such as hart’s-tongue, nipplewort, wood meadow-grass and shining crane’s-bill. Others must have spread from dry grassland on limestone or heaths, such as red fescue, early hair-grass, sticky mouse-ear and common mouse-ear. Many of the occasional wall species are normally found on arable or disturbed ground, as witness Kilvert’s poppies, but also shepherd’s-purse, annual meadow-grass and squirreltail fescue.
Curiously, many native species are far more abundant on walls than in their natural habitats, which implies that they increased when people started to build with stone. The rustyback fern, for example, often covers limestone walls, but is rarely found on rocks. On the other hand, by no means all the species of rock faces, dry grassland and wooded rocky slopes have colonised walls. Species such as dwarf sedge, bloody crane’s-bill and horseshoe vetch have remained resolutely on their natural fastnesses, and many uncommon and rare species of quarry spoil (e.g. yellow-wort), wooded screes (e.g. narrow-leaved bitter-cress), dry margins (e.g. madder) or dry heaths (e.g. wood sage) show little or no sign of spreading to walls. The polypodies seem equally common on walls and natural habitats, including the rare southern polypody, which has been found on both Goodrich Castle and at King Arthur’s Cave.
QUARRIES AND PITS
People have been making holes in the ground for as long as we have built in stone, surfaced roads and spread lime on sour land. Until the nineteenth century, quarries and pits in the Lower Wye were small, though rock towers near the Longstone were ‘gormandized’ by the ‘ravenous lime-kiln’ at New Weir (Hall, 1861). Many were hidden in woodland, where they were eventually recolonised by trees. Latterly, however, quarries have become few but extremely large. In fact, the cliffs and screes produced are larger than those that were created naturally.
Active quarries have little vegetation, but they vegetate once they are abandoned. Spoil heaps rapidly develop a tall-herb assemblage that becomes grassland if rabbits, deer or sheep have access. Quarry floods, with impeded drainage, develop patches of marsh, and faces slowly acquire cliff species. However, quarries in woodland need to be fairly large if they are not to be shaded by overhanging trees. When the scowles, such as those in Dingle Wood, were mined, coppice stools were left on the tops of rock pillars and margins, so they have become little more than gloomy, leaf-filled chasms with a scatter of hart’s-tongue and other ferns (Fig. 143).
A particularly interesting quarry under Caplar Hill was visited by the Woolhope Club in 1883 (TWNFC, 1883-5, p. 52). This was once a natural landslip, but it had also been quarried for stone to build Hereford Cathedral. ‘On approaching the quarry from the river the appearance of yellow-wort quickens the botanist’s attention, and shows him that he is in a place where some “good things” may turn up.’ In fact they found pale St John’s-wort, bee orchid, wayfaring-tree and fingered sedge – a respectable haul, though the orchid was mistakenly given the scientific name of the fly orchid, a much rarer species.
These days, the smaller, older quarries have lost most of their former
FIG 143. Deep scowles within the settlement of Scowles. Almost all the many tracts of old mine workings are hidden in ancient woodland of beech, yew, ash and other species.
specialised biological interest. Their fate is exemplified by the ‘Great Quarry’ on the Doward, which was quarried for lime and road metal until the early twentieth century. When Eleanora Armitage (1914) knew it, the worked-out face and floor were partly overgrown with shrubs and bushes, but the ground was largely covered ‘by a rough stony turf plentifully sprinkled with a large number of flowering plants and mosses’. The dominant grasses were false brome, tufted hair-grass and common couch, with mountain melick, quaking-grass, fern-grass and sheep’s-fescue, forming a turf with many characteristic species of limestone grassland and scrub, such as fairy flax, salad burnet, wild carrot, blue fleabane, yellow-wort, autumn gentian, bee orchid and fly orchid. Nearly a century later, the same quarry is overgrown with woodland. A rim of beech-yew woodland shades the quarry faces and a tall stand of mixed beech-ash-wych elm-maple-large-leaved lime casts a lighter shade on the floor. Two large old beeches stand in the quarry floor, one of which has collapsed, leaving a clearing containing a few moderately interesting species, such as deadly nightshade and stinking hellebore. Some of the original grassland species, including glaucous sedge, red fescue, salad burnet, yellow-wort, fairy flax, common milkwort, corn mint and madder, survive nearby in a small glade beside a track, and two fine greater butterfly-orchids hide in the fringing scrub.
Quarries and pits acquire their plants from surrounding vegetation. The older pits, having formed in an environment rich in semi-natural vegetation, provided refuges for sometimes-rare heathland, grassland and woodland species that subsequently vanished from their surroundings. No doubt this is how the fly orchid colonised the debris of other lime quarries on the Doward (Purchas & Ley, 1889), and how autumn lady’s-tresses came to be found in a cornstone quarry near Garway Hill by the Woolhope Club in August 1954. However, their function as refuges ends when quarries fill with scrub and woodland. Thus the gravel pits within the conifer plantations of Trellech Common, which have heathland species such as heather, wavy hair-grass, trailing St John’s-wort, bell heather, changing forget-me-not and wood sage, and a seasonal pool with floating sweet-grass, yellow iris, soft-rush, sharp-flowered rush, pondweeds, greater spearwort, wood club-rush and bulrush, will retain these species only if it is kept free of woodland. Likewise, the miniature quarry faces of road cuttings develop rich vegetation only while they remain at least partly open to full sunlight.
The modern super-quarries are concentrated in the Carboniferous Limestone. They may eventually develop into rich habitats, but not if well-meaning attempts to ‘restore’ spoil heaps with tree planting are continued. This politically correct practice shortens the species-rich open and scrub phases of succession and prevents birch, sallow and ash woodland from developing
FIG 144. Pen Moel cliffs and quarry as seen from the Wyndcliff, with the tidal Wye just visible below. The abandoned quarry and screes of residue have reverted to woodland, though the mix of species differs from that of the scraps of ancient woodland that survived the quarrying. The houses of Woodcroft extend to the cliff edge. Beyond are the fields of Sedbury Park and the rim of ancient woods that separates them from the Severn estuary.
naturally. Where natural succession has been permitted, as it has on the spoil heaps and screes left by the great quarry at Pen Moel cliffs near Chepstow (Fig. 144), species-rich vegetation has developed, though it is now proving to be a point of entry for non-native species, such as the sycamore, buddleia and cotoneaster.
GARDENS AND OTHER DISTURBED GROUND
Villages, farms and isolated houses provide distinctive habitats characterised by disturbance (Fig. 145). Farms and woods are full of bonfire sites, trampled gateways, rutted tracks and other small patches of disturbed ground that develop a specialised vegetation, such as the fireweed and Funaria hygrometrica of bonfire sites and the plantains and scentless mayweed of farm gates. Farmyards, themselves, are a kind of disturbed patch, where nettles and docks are common, and swallows and barn owls once thrived. However, the most extensive disturbed habitats after arable fields are gardens and industrial wasteland.
FIG 145. The church at Goodrich with the house once occupied by Joshua Cristall (lower left) (see also Fig. 111). The mixture of gardens, rough ground, churchyard, semi-natural grassland, scattered trees and walls that characterises most villages provides far greater habitat variety than the commercial farmland nearby.
Gardens have been a feature of the Lower Wye for at least 2,000 years (Whittle, 1992), for both the legionary fortress at Caerleon and the courtyard houses at Caerwent had enclosed gardens. In 1270, Chepstow Castle seems to have included a garden within the walls. Tintern Abbey not only had two gardens at the abbey itself, but also a garden at Rogerstone and a walled garden at Merthyr Gerain. In about 1445, Raglan Castle had orchards full of apple trees and plums, figs, cherries, grapes, French plums, pears and nuts, and the remains of fine Tudor and Stuart gardens can still be seen there. In the early sixteenth century, gardeners from Troy House went to the Continent and returned with plants and an awareness of latest garden fashions. In 1686, Jacob Millard’s map of Chepstow shows numerous town gardens and orchards.
Gardens provide a fine-scale mosaic of grassland, disturbed habitats and semi-woodland. Lawns may be perfectly good examples of semi-natural grassland if the householder applies no fertiliser, herbicide or muscicide (Chapter 6). Gardens close to ancient woodland constantly remind us of the power of natural habitat restoration: ours and several other gardens on the fringes of the woods lining the lower gorge are constantly invaded by bluebells and stitchwort. ‘Gardens of remembrance’ (Figs 146 & 147) have latterly been recognised as
FIG 146. Ivy broomrape in the churchyard at St Mary’s, Chepstow, formerly the church of a Benedictine priory. The churchyard supports a vigorous population around gravestones and in shady margins.
FIG 147. Wild daffodils among the graves in the churchyard at St Mary’s, Kentchurch, an oasis of semi-natural grassland in an intensively cultivated neighbourhood. The principal lichens on the gravestones were identified by Alan Orange as Ochrolechia parella and Tephromela atra, with scraps of Melaneliaglabratula, Candelariella vitellina and Caloplaca holocarpa.
important wildlife habitats with reasonably good examples of semi-natural grassland, refuges for uncommon species and excellent facsimiles of small cliffs.
Most borders and vegetable gardens are, of course, largely stocked with foreign species, though the present interest in wild gardening has allowed some colourful arable weeds, such as corncockle and cornflower, to regain a foothold. Every gardener knows that there are several attractive plants that are virtually self-sustaining in gardens. We, for example, have no difficulty maintaining caper spurge, celandine saxifrage, green alkanet and dame’s-violet in this fashion, and our stream is now graced with monkey musk, washed down from a garden up the lane. Many of these escape briefly into the wild, and some become entirely naturalised, a process that must have started at least 2,000 years ago.
The less attractive wild garden species are ‘weeds’, of which ground-elder, bindweeds, creeping buttercup, dandelion, nipplewort and couch grass are among the most virulent. No doubt the composition changes, but most have been around for a long time. Purchas and Ley (1889) noted that creeping buttercup was then a ‘troublesome weed’ in Herefordshire gardens; that ground-elder was most frequent in garden hedges (‘a very troublesome weed, from the deep running and brittle roots, the least piece of which will grow’) and that petty spurge was very common as a garden weed. From whence did they spring? Some clearly emerged from disturbance patches in natural habitats (e.g. nipplewort from tree-fall gaps in woodland), others are constituents of semi-natural grassland (e.g. couch grass), but a significant minority appear to have come from riversides. Ground-elder, bindweeds, creeping buttercup and stinging nettles grow in the moist, well-fertilised, lightly grazed and moderately disturbed conditions of natural floodplain forest and flushed woodland, so it is hardly surprising that they like gardens.
WOODLAND RIDES
Rides and forest roads occupy 5-10% of a woodland’s area (Fig. 148). Structurally, they consist of the central bare road surface or turf kept short by mowing and traffic, verges of taller grassland and edges defined by ditches choked with tall herbs. Shrubs behind the ditch form a boundary zone that grades back into the trees. When both the adjacent stands are mature the ride vegetation is somewhat shaded, but when at least one side is young, the whole ride receives almost as much light as open fields, at which time the ‘carriageway’ is roughly analogous to pasture, the margins to meadow and the ditches to marsh. Rides and roads are designed for access and timber extraction, so they are regularly disturbed. Bare
FIG 148. A forest road within oak-dominated woodland in Haugh Wood, opened and structured by the Forestry Commission with fritillary butterflies in mind.
road surfaces admit many ruderal species, while poorly drained rides are sometimes left with ruts that develop into marsh, and may hold water long enough to entice frogs to spawn. Ride orientation also influences habitats: at the extremes, east-west rides have a sunny, hot, dry northern margin and a shaded, cool, moist southern margin.
The character of ride vegetation depends on how the wood originated. Rides in secondary woods preserve modified remnants of the habitats that preceded the woodland, whereas the species found in ancient-woodland rides probably colonised from outside the wood. Preservation of precursor habitats in secondary woods is best exemplified by the heathland remnants in the plantations on the former Wyeswood Common and Tidenham Chase (Chapter 6) and survivals from grass-heath (e.g. climbing corydalis and heath speedwell) in Ninewells Wood and Beacon Hill. Equally, wild liquorice in plantations near Woolhope and adder’s-tongue and cowslips in the shade of deciduous plantations testify to the former presence of semi-natural grassland. Another possibility is that some species have survived on rides since an earlier wood-pasture phase: this may lie behind the presence of small patches of heather on ride margins in the ancient Chepstow Park Wood and Haugh Wood and even some of the grassland species in other ancient woods. However, the capacity for rides to act as refuges for heathland and unimproved grassland is strictly limited: the common species may survive, but the specialists usually vanish.
In ancient woods the tall herb margins take the form of meadows, especially where the foresters mow them in late summer. In terms of the National Vegetation Classification they fall close to MG1, false-oat grassland, which is typical of ungrazed and unmown verge and field margins, particularly the sub-communities characterised by meadowsweet or knapweed (Rodwell, 1992). Some of the commonest species are meadowsweet, great willowherb, marsh thistle, perforate St John’s-wort and slender St John’s-wort, embellished with fine displays of common spotted-orchid. Rides on limestone (Fig. 149) may be
FIG 149. Wild columbines by the forest road in the Minnets, on the limestone between Caerwent and Rogiet. Once one of the richest woods in southeast Monmouthshire, but now largely converted to even-aged plantations, the rides and roadside have remained open, and still support a rich flora.
graced by columbine, early-purple orchid and, towards the Severn, upright spurge. Shorter turf often includes selfheal and common bird’s-foot-trefoil in abundance, together with characteristic limestone species such as fairy flax, common centaury and yellow-wort. Sedges, such as hairy sedge and yellow-sedge, are well represented on wet ground. Shaded rides often have much thin-spiked wood-sedge and thickets of pendulous sedge. Ride vegetation is not a true meadow, however, for good meadow species such as yellow-rattle are rare, and shade species, such as wood anemone, invariably infiltrate.
Rides and other open spaces contribute half of all the plant species in a large wood (Peterken & Francis, 1999), and they are even more important for butterflies and other insects (Chapter 10), so woods are much impoverished when rides become shaded. Trevor Evans is sure that the ride flora of the limestone woods around Mounton and elsewhere has declined, due to both neglect of coppicing and the subsequent replanting with beech over a short period. Even small changes have had an impact. For example, uncommon grassland species, such as pyramidal orchid, once survived in woodland entrances, but they have now been eliminated by scrub growth. Ironically, the richest floras may well be in rides that were kept open by conifer planting and harvesting. Rides are also proving to be points of entry for alien species, such as Indian balsam, which is infiltrating woods at Bigsweir along newly created forest roads.
HEDGES
Wordsworth described the hedges above Tintern as ‘little lines of sportive wood run wild’, but one wonders how wild they really were. Hedges were made as highly functional stock barriers and property boundary markers, and they had to stay in trim to be useful. Some originated as boundaries of woods that were subsequently cleared, others by natural seeding along boundaries or by planting. Some originated before the Normans arrived (Chapter 4), but others were formed later when common land was enclosed and larger fields were divided. A few are very recent indeed, planted around fields under stewardship agreements and along realigned roads. In general, older hedges have the richest flora and greatest diversity of shrub species, partly because they sometimes contain trees and shrubs, like small-leaved lime and spindle, which are very slow to colonise. Older hedges tend to be sinuous, whereas the hedges of planned enclosure tend to be straight.
Hedges are not just lines of shrubs containing a scatter of trees (Fig. 150). Most grow on banks and thus have drier soils than their surroundings, and many
FIG 150. Thick, well-treed hedges forming the field boundaries at Springdale Farm, and linking to ancient semi-natural woodland to the left. Located on the extremity of the Trellech-Wentwood plateau close to the Usk valley, this is now a Gwent Wildlife Trust reserve where the semi-natural grassland is being maintained as meadows (recently cut) with aftermath grazing.
are associated with a ditch. At the base of the hedge is unploughed, unmown and less-grazed herbage that German ecologists recognise as a distinctive border community, or zaum. In the Lower Wye the great majority of trees are oaks and ash: indeed, southern Herefordshire is still notable for a high density of fine spreading oaks, some of which are pollards, and a few large beeches remain from what seems to have been a late-eighteenth-century planting fashion. In and around the gorge, great limes are still frequent, and many of these have been pollarded, some more than 5m from the ground. In the past, hedges were carefully cut and laid, and this continues on a small scale. Today, a remarkably high proportion is carefully trimmed.
Rebecca Roseff’s (2003) study of the hedges in Marden, in the lower Lugg valley, tells us how composition changes with age. Most nineteenth-century hedges were planted with hawthorn, and have since been colonised by elder, ash, oak and, on former marshes, crack-willow. Earlier post-medieval hedges were planted with hawthorn and English elm. Earlier hedges, most of which bordered ancient lanes and manor boundaries, generally have far more hazel and field maple than hawthorn, but they lack woodland herbs, such as dog’s mercury. Their age as hedges was difficult to determine, partly because medieval references could mean dead-hedges, not live ones. No doubt their limited range of species reflects the fact that Marden has been a largely open landscape for 4,000 years.
Marden is not representative of the Lower Wye, and indeed each district probably has its particular hedge-history and combination of species. Most Lower Wye hedges are mixtures, which in theory indicates that they are several hundred years old. At first sight most seem to be combinations of hawthorn, hazel, bramble, rose and honeysuckle, but closer inspection reveals great variety. This is partly explained by soils: gorse appears in hedges on light, dry soils, whereas spindle and field maple are more prominent on heavy and lime-rich soils. However, hedge origins are also significant. For example, some around the gorge formed over the last 200 years as self-sown lines of scrub along low walls, and these tend to be dominated by blackthorn, hawthorn, elder and ash, with sallow on wet and heavy soils. Others must have been planted as mixtures – for example, a hedge on St Briavels Common that did not exist in 1927, but which is now a line of cherry plums, self-sown oak, beech and sallow, and planted limes. Since the plums have the full range of fruit colours from deep purple to yellow, with shapes ranging from ovoid to spherical, one assumes that they were planted for variety. The English elm-dominated hedges north of Monmouth also appear to have been planted.
Most hedges, however, must be much older, since they define ancient boundaries and comprise mixtures of hazel and holly among the thorns, and a scatter of small-leaved lime, maple, wych elm, beech, dogwood, spindle, guelder-rose, wayfaring-tree and others (Fig. 151). Some, such as the overgrown hedges of the Coxbury Lane in Newland, are certainly wood-relict hedges, i.e. ancient woodland boundaries that were kept as hedges when the wood was cleared. Their original condition can still be understood by looking at original boundaries round ancient woods: although invariably overgrown, they still have horizontal trunks running along the boundary bank, a sure indication of past cutting and laying. Other, equally ancient and diverse hedges could have been planted with transplants from the nearest wood, but Rebecca Roseff has found no records of ancient planting in Herefordshire. Certainly, the hedges of the Monmouthshire back lanes look as if they have been carved from woodland centuries ago.
How long do hedgerow shrubs live? Some of the limes and wych elms may be potentially immortal – youngish growth maintained on ancient rootstocks by traditional cutting and laying (a form of coppicing) after clearance of ancient woods a millennium or more ago. Clearly, there is some death and renewal
FIG 151. A mown hedge near St Briavels, with semi-natural MG5 grassland behind. This hedge appears to be ordinary, but it comprises a mixture of tree and shrub species, including small-leaved lime. The lime is a frequent constituent of Lower Wye hedges, probably because many were formed directly from ancient woods.
within a hedge, for sycamore and other introductions have displaced other shrubs in field and road hedges near houses. Today, most hedges are mechanically trimmed, so hedge bottoms become hollow, and some side branches, ramifying through the thicket, sometimes take root many metres from the parent shrub.
Many woodland herbs can be found in hedge bottoms, but there is a bias to the species of dry, alkaline woodland – such as dog’s mercury, wild arum, bluebell, bracken, primrose, wild strawberry, wood mellick and common dog-violet – and wood-margin species such as greater stichwort, red campion, black bryony and even madder on the limestone around Mounton. Wood-relict hedges elsewhere may retain slow-colonising woodland herbs, and such species colonise readily enough close to ancient woods. Likewise, some of the roadside hedges growing on former heathland still contain bilberry, but mixed with the lime-loving species mentioned above, ajuxtaposition that would not be found in woods, but which has presumably come about by the invasion of the woodland herbs onto sites enriched by subsoil disturbance. Hedges in much of the Archenfield area of Herefordshire have poor woodland floras, but, as Rebecca Roseff tells me, this area was described in several Llandaff charters as an unhedged open landscape, used principally as extensive pasture.
Two modern factors have undoubtedly had a great effect on the hedge flora. In the 1960s, Monmouthshire County Council started a disgraceful habit of controlling ‘weed’ growth with herbicide sprays. When I first knew the district the Monmouthshire hedge bottoms were extremely colourful, but spraying devastated the primroses and, oddly, the blue form of sweet violet. The verges were also sprayed: as one resident recalled, ‘before the Council sprayed the verges from the Whitebrook crossroads to the Peckett Stone, there were win berries, orchids, cowslips, ferns and many wild flowers’ (Bevan & Rees, 2005). The spraying has stopped, but the flora has not recovered. Its effects can still be understood in minor lanes, such as those round Kilgwrrwg Common, which presumably escaped. Recent decades have also seen the widespread application of fertilisers and herbicides in agriculture, with the result that hedges contain a great deal of nettles, cow parsley, cleavers, hogweed and other species that respond to eutrophication.
ROADS, TRACKS AND GREEN LANES
These are the quintessential linear habitats. One does not have to believe in Alfred Watkins’ Old Straight Track to recognise that roads may be the oldest habitat created by people, and that some may have been derived from tracks created by wild cattle and deer. The similarities between the Roman road network and modern routes (Chapter 4), combined with continuity of settlements, imply that some modern roads and tracks have existed for thousands of years. On the other hand, the parish summaries in the Victoria County History (Herbert, 1996) show that the track and road network was constantly evolving: not only were new routes constructed, but routes that had been important could fall into disuse. Heather Hurley’s (2000) account of green lanes in Herefordshire confirms that many of today’s minor tracks are ancient and formerly well used.
Roads were not sealed before the modern era. Those on slopes stayed dry, but sunkenways down steep slopes ran like rivers after rain, and where tracks crossed poorly drained soils and dips in the ground they could become impassably wet and rutted. Even in the era of the Wye Tour, the roads of Monmouthshire were ‘capable of much improvement; in some parts they are infamously bad’ (Fox, 1794). Nevertheless, metalled surfaces are not entirely new, for the Romans paved their main roads, and the Cistercians certainly took some care of the roads between their abbeys and farming properties (Williams, 2001, p. 194). Unpaved greenways connected Grace Dieu with its Penrhos grange, but a medieval road sweeper is recorded at Merthyr Gerain, and Tintern had paved the road from Brockweir to Tintern ferry (currently one of the wettest green lanes in the Wye Valley!) and its ‘stony way’ from the abbey gatehouse to its vill at Porthcaseg.
Tracks provide a central carriageway of ruderal plants and, when rutted, small marshes, but their key habitats are the grass verges. Roads have a history of grazing: most were grazed by stock being driven from one field to another, some were mown for hay, and lanes crossing former commons retain verges derived from the semi-natural grassland of the common pastures. With this background, one might expect verge grassland to be as rich as the grasslands of pastures, meadows and commons, but there is little evidence of this in the old Floras. True, there are limited indications that verges were distinctive, and certainly they would have been more sheltered and disturbed than other grassland. Thus, upright hedge-parsley was strongly associated with verges, and some introduced species, such as greater celandine and dame’s-violet, may have spread best along roads. In the Cotswolds, meadow crane’s-bill was abundant on verges, but it is infrequent in the Lower Wye, and seems to be more associated with river banks.
Today, commonable sheep still graze the fertile verges of the Dean in preference to the less fertile grassland among the trees, but otherwise there is a strong suspicion that verge floras have deteriorated in recent decades. Monmouthshire’s roadside verge cowslips and spotted-orchids were decimated in the late 1960s and 1970s by herbicides, and they have not recovered since the spraying stopped. The combination of infrequent mowing and fertiliser drift from adjacent fields has allowed a few vigorous species (e.g. cock’s-foot, false oat-grass, hogweed and cow parsley) and disturbance beneficiaries (e.g. dandelion, broad-leaved dock) to dominate and oust the herbs of smaller stature. Nevertheless, in districts used mainly as arable and sheep pasture, the verges remain refuges for a few semi-natural grassland species, notably cowslip and cuckooflower, and there are still some gems on little-used country lanes (Fig. 152). The Ross spur motorway verges are famous for their wild daffodils (Chapter 9), and pyramidal orchids grow within a metre of the speeding traffic south of Monmouth. The salt-tolerant Danish scurvygrass has spread along main roads.
FIG 152. Early-purple orchids resplendent on a lane verge near Aconbury Court. Road verges were important refuges for wild plant species until recently, but most have been so enriched by drifting fertilisers and swamped by rank growth that is no longer grazed or cut for hay that they now hold little more than widespread, vigorous species like cow parsley.
RAILWAYS
In 1810 a horse-drawn tramway was built from Monmouth to Coleford and the collieries beyond, and a year later another was opened between Abergavenny and Hereford. Railways arrived in the mid nineteenth century: the Severnside line through Chepstow in 1850, the Hereford-Ross line in 1853, Ross-Monmouth in 1873, and Chepstow-Monmouth in 1876. The lines in and around the gorge all remained quintessential branch lines, where staff transported chickens and knew passengers personally. They finally closed in 1964-5 (Handley & Dingwall, 1998; Jenkins, 2002).
The lines generated new habitats – cinder-tracks, embankments, cuttings and grassland maintained by rabbits and accidental fires. When constructed, they were evidently responsible for the revival of henbane and corn marigold from buried seed (TWNFC, 1852-65, p. 142). At Chepstow sidings, pale and common toadflax met, fell in love and produced an array of hybrids. Presumably, dry grassland developed on embankments, ruderal assemblages on the track and small patches of wall and rock-face vegetation in cuttings, but few observations survive to confirm this.
Today, where they have not been re-incorporated into neighbouring fields or reclaimed by adjacent woodland, the abandoned lines have become footpaths, forest roads and inaccessible lines of hawthorn and rank grass, leaving only a limited legacy of distinctive vegetation. One diverse example is the track from Tidenham tunnel to Tintern, which is kept open as a walking track and forest road where wood vetch does well on the dry limestone banks and the remote end of the track bed is regenerating to small-leaved lime woodland. Most tunnels have been bricked up, leaving openings for the use of bats, but the Tidenham tunnel remains open. The operational lines along Severnside are not accessible to biological survey, but southern polypody is well established below Bulwark. The recent ‘coppicing’ above the tracks at Gatcombe and Chepstow is known to have felled several rare whitebeams.
LINEAR HABITATS, SPECIES MOVEMENT AND ECOLOGICAL ISOLATION
Now that we have catalogued the main linear and patch habitats, we can consider whether they have in fact helped species to move around the landscape. This is important, for movement enables species to expand populations, colonise new habitats, repopulate habitats from which they have been lost during, say, a period of neglect, and refresh the genetic diversity of populations in otherwise isolated habitats.
Species certainly move along rivers, and not just the fish and other aquatic species that have no alternative, but also woodland plants (Chapter 7). Purchas and Ley (1889) described how chives, alpine enchanter’s-nightshade [sic], sheep’s-bit and wood stitchwort had been washed down from higher up the Wye to grow in lowland Herefordshire, and in recent times we have witnessed the spread of Indian balsam along rivers. Elsewhere, beetles and other invertebrates have been shown to move along hedges, and bats are known to forage along valleys, partly because this is where their preferred mix of pasture and trees is most likely to be found, and they use overgrown hedges as safe routes between roosts and feeding grounds. The classic study of plant movement in an agricultural landscape by Woodroffe-Peacock (1918) found that many plants spread along hedges and tracks, and that others were carried as seed on muddy vehicles and in trouser turn-ups. Herefordshire’s dormice reach woods better if they have hedges along which to move (Chapter 10).
However, there is plenty of evidence that linear habitats make only a modest contribution to wildlife movement. Most hedges contain only a limited range of woodland plants, which implies that few use them for migration. Furthermore, some woodland species can cross unsuitable habitats, such as the great spotted woodpeckers we see flying across a field to perch in a hedgerow oak, before going on to the next wood. In this instance, the tree is a staging post or stepping stone between woods, so it hardly matters that it is part of a hedge. Grey squirrels, too, will cross grassland, but they depend on a quick dash for safety and must prefer hedge-trees to isolated trees. Badgers and foxes are just as dependent as squirrels on woods and trees, but happily cross fields and pass through, not along, hedges. It seems, then, that the distances between woods, and the density of stepping stones in farmland, may be just as critical for wild species as the uninterrupted hedges.
For some species linear habitats are no use at all. Consider the predicament of the green-winged orchid, a species that was widespread in the meadows and pastures of traditional mixed farms, but which has been reduced by ploughing and reseeding to a few scattered populations. One of these populations is in a small field at Penallt, owned and managed as a meadow by Geoffrey and Vera Easton. One day, however, they will no longer be the owners, and the future owners may neglect or reseed the field, thereby exterminating the orchid. If, subsequently, the field were again treated as a meadow, the sward would eventually revert to a mixture of native grasses and flowers, but we could not be sure that the orchid would return. It would have to come back from a population that had survived elsewhere, and this is unlikely unless the nearest population is nearby (say, in the adjacent field) and/or large (produces a great deal of seed). In the early twentieth century, when semi-natural grassland was still a matrix habitat, the old Floras record that green-winged orchid was common, so any field that became a meadow stood a good chance of being colonised, but today, the orchid populations are few, small and far between, and the chance of them colonising is much reduced, however suitable the site may be. Orchids cannot colonise from field to field in short jumps because most fields are ploughed frequently, but they cannot move stealthily along hedges and roadside verges because competition from stronger species would never let them get established. So they would have to rely on seed from several kilometres away drifting to exactly the right spot, a possible but supremely unlikely event. Thus, the green-veined orchid population of this meadow is isolated. More generally, the Lower Wye population of green-winged orchids has been fragmented into a thin scatter of separate populations, between which there is no communication, each of which must sur vive in isolation. Only new meadows that lie close to existing populations stand much chance of being colonised.
These considerations apply with particular force to woodland, the original matrix habitat, which has been fragmented for millennia. Woodland species evolved in an environment where movement was barely necessary and movement across the landscape was unrestricted by hostile territory. Any ground that reverted to woodland after a disturbance would be readily colonised by woodland species, which were abundant and nearby. Under natural conditions, many woodland plants produce little fertile seed, disperse only over short distances (by ants), or spread vegetatively by runners or rhizomes, but this was no disadvantage. Indeed, when new woodland develops next to ancient woodland, the woodland plant species move in at about 1m per year (Bossuyt et al., 1999). However, as prehistoric forest became ever more fragmented into separate woods, so woodland species found themselves ill-adapted to the man-made landscape, and it is therefore no surprise to find that at least 40% of the woodland flora is slow to colonise from ancient to new woodland, except when the new wood is close to an ancient wood (Peterken & Game, 1984).
Woodland plants spreading at 1m each year would spread 1km in 1,000 years – and would not yet have returned to Britain after the last glaciation. Fortunately, rare instances of long-distance dispersal (e.g. seed or rhizomes might be washed some distance down a stream) must have helped them to move faster, and we see signs that this is what happens in practice from the diversity of woodland plants along streams (Chapter 7). Today, however, we have limited the opportunities for inherently slow-moving species to spread by isolating woods within intensively used farmland. Species that live in arable and leys – today’s matrix habitats – are not isolated, but for many, perhaps most, of the species in semi-natural habitats isolation is a factor in their long-term survivial.
Can we express isolation precisely? Is there a threshold area below which a habitat becomes subject to isolation effects? Remarkably, theory and studies of individual species both point to a key threshold around 30% cover (Peterken, 2000), though this varies according to the shape of the habitat patches. If woodland, for example, occupies only 5% of the land, remaining woods will tend to be small and new woodland will tend to be isolated from existing woodland. As the woodland area increases, so woods become larger and closer together, until at about 30% cover we reach a point where many woods are large, and any new woodland will necessarily be close to an existing wood, so that it can be colonised even by species that move only in short jumps.
Using this as a yardstick, we can divide the Lower Wye into two kinds of landscape. In and around the gorge, woodland covers more than 30% of the land, and has always done so, and many woods are large (Chapter 3). Moreover, the land between the woods has numerous thick hedges, green lanes, boundary trees and semi-natural grasslands, so the ‘matrix’ is benign, i.e. woodland species have no difficulty moving through it (Fig. 153). Semi-natural pastures and meadows have limited cover, but they are to some extent linked and buffered by rides within woodland. Under these circumstances we should expect woodland species, and to a lesser degree grassland species, to be resilient, and indeed bluebells, primroses and wood anemones, for example, are common in field boundaries and quickly colonise new woodland. Similar circumstances exist in the Dean, on the Woolhope Dome, and locally on the fringes of the Trellech plateau and the southeast Monmouthshire limestone. On St Briavels and Hewelsfield commons, semi-natural grassland covers well over 30% of the land, and our experience is that the flora of fields recovers fairly well from periods of cultivation or intensive grazing.
Elsewhere, woods and semi-natural grassland cover a much smaller proportion of the land. Although some are large, they are isolated, and the farmland between offers few opportunities for species to spread. True to expectation, bluebells, primroses and wood anemones are infrequent in hedges of the lightly wooded parts of south Herefordshire, except in the vicinity of woods and ancient semi-woodland habitats, such as sunken, shaded stretches of green lanes and minor roads, where they are common.
Until the twentieth century, farmland was a sub-equal mixture of ploughland and semi-natural grassland. Both land types occupied 30% or more of the
FIG 153. A typical hedge within the Hudnalls, full of bluebells and other woodland plants in a matrix of semi-natural grassland. In this small-field landscape, woodland species are all-pervasive.
landscape; mixed farming ensured that they were evenly distributed; hedges and streams formed an intact network of links; and under these circumstances isolation was limited. Today, isolation is much increased, for a number of reasons: much more farmland is arable; most grassland is sown or ‘improved’; the network of hedges has been torn; woodland habitats are less diverse; and most of the landscape is awash with excess fertiliser (which limits the number of species that can compete in farmland habitats). The quarries, pits, railways, roads and gardens have provided limited respite, especially those that go close to, but do not destroy, semi-natural habitats, but they have rarely proved enduring as refuges.
Ecological isolation is a natural factor. In the Lower Wye, for example, the plants of limestone outcrops must always have been isolated. For other species, such as birds, dragonflies, dandelions and inherently fast colonists, it may be irrelevant. For many species, however, isolation is an important limitation, so the considerable increase in ecological isolation during the twentieth century (even in the Lower Wye!) presents a problem. ‘Divided they fall’: the conservation challenge is to manage the landscape so that ‘together they can stand’.